Is it possible to update CK2's engine?

[Dsingis]

I'm no programmer, and I'm no game designer, so if this is a stupid question, excuse me

But, I've heard the game devs say, that CK2's performance issues, and instability comes from the engine being so overloaded with features, which was never intended.


Now, I was asking myself, is it theoretically possible to update CK2's engine to make it able to handle so many things easier, without basically programming CK3?

I mean IF it was possible, I think many players here would agree that rather than a new content update, we would love to have an engine update DLC.

But, like I said, I'm no game dev, so I'm asking, is this even possible?

Cheers!

 

[HandicapdHippo]

Engines aren't static things, especially ones created by the devs themselves like Paradox's engine is, CK2 and its engine has been contentiously improved by the developers the whole time and even had improvements developed by the teams working on other paradox games ported over to it. A lot of the optimizations that came with the Reaper Due patch and other patch's are basically exactly what you are asking about.

 

[Rags17]

The game as it stands now includes hundreds if not thousands of events and decisions that only appear if a DLC is active, or is inactive or whatever. The game still has to check these things every single cycle, so if these were all specifically included or excluded then this alone would save a good chunk of processing time.

There is also lots of "orphan code" that is no longer used but which was necessary for earlier versions and or earlier patches. Things like old style assassinations, old versions of landed titles (eg Papacy, Caliphate) etc.

Finally, all of the game's events and decisions use their own unique custom flags, modifiers and opinion_modifiers. These could be cleaned up and amalgamated to streamline event handling and minimize clashes and weird outcomes. For example, there must be half a dozen opinion modifiers that all amount to "Disappointed" "Angry", or "Upset", all of which have the same or similar modifier effects.

In short - even without changing the engine I am certain that there are lots of ways the game could still be optimized further.

 

[klopkr]

Paradox updates the clawitzengine all the time on the fly. Usually they do it between games though.

Can you name one thing they can add to the engine to make the game run faster? I'm pretty sure they prefer that much more than just 'make game better'.

 

[Rags17]

Paradox updates the Clausewitz engine all the time on the fly. Usually they do it between games though.

Can you name one thing they can add to the engine to make the game run faster? I'm pretty sure they prefer that much more than just 'make game better'.

The ability to "turn off" great chunks of the map. This alone would yield a HUGE improvement in speed.

The mechanic is relatively straightforward too - simply add a (player inaccessible) routine that turned all "switched off" provinces into unplayable terrain and remove all holdings from these spots. The instant permanent destruction of 20-50-90% of all characters in game and suddenly the game will FLY.

I think that the problem is that game as it stands now WANTS to keep repopulating courts all over the map - the trick would be to add the functionality to tell it so not do this for the targeted areas. It seems straightforward to me but what do I know, there must be something making it hard to implement.

 

[klopkr]

The ability to "turn off" great chunks of the map. This alone would yield a HUGE improvement in speed.

The mechanic is relatively straightforward too - simply add a (player inaccessible) routine that turned all "switched off" provinces into unplayable terrain and remove all holdings from these spots. The instant permanent destruction of 20-50-90% of all characters in game and suddenly the game will FLY.

I think that the problem is that game as it stands now WANTS to keep repopulating courts all over the map - the trick would be to add the functionality to tell it so not do this for the targeted areas. It seems straightforward to me but what do I know, there must be something making it hard to implement.

This is already possible in two ways.

The first is just simply making all the lands you don't like into wasteland and the second is simply using a mod or dlc that changes the map. The second one is actually literally what happens when you go back some patches. It has nothing to do with the game engine.

The ability to change the map on the fly would be nice but would require immense work for a tiny feature as the map is loaded on game start up and not in game. To change it would ether move the loading screen to in game which would remove the players ability to just pick a start on the map or double the loading screens. I don't think anyone wants to do this for you to remove a region that hasn't effected game speed since a few patches ago.

Regardless of all of that Pdox has said many times they want to avoid anymore 'versions' of the game to update. Keeping the game balanced without a certain area is significantly more work to make a very small community happy through what is pretty much a placebo. They rely on modders and old patches to do that for free.

Name something else that could be added to the engine to improve the game.

 

[Foreign Alien]

I think what CK3 engine needs is a way to have modular map, so that they can expand it however they want without encumbering those who don't want the new stuff.

 

[Andrelvis]

I'm no programmer, and I'm no game designer, so if this is a stupid question, excuse me

But, I've heard the game devs say, that CK2's performance issues, and instability comes from the engine being so overloaded with features, which was never intended.


Now, I was asking myself, is it theoretically possible to update CK2's engine to make it able to handle so many things easier, without basically programming CK3?

I mean IF it was possible, I think many players here would agree that rather than a new content update, we would love to have an engine update DLC.

But, like I said, I'm no game dev, so I'm asking, is this even possible?

Cheers!

It is possible, but to change fundamental parts of the code would take a lot of work. And quite significantly, it would likely need extensive QA. So it should be possible, but at the same would require a significant investment from Paradox. It depends if they think investing in the game for the long term is worth it or not; or if financially it would be better for them to make CK3 instead.

 

[Andrelvis]

I think what CK3 engine needs is a way to have modular map, so that they can expand it however they want without encumbering those who don't want the new stuff.

That would be great indeed. My computer wouldn't be able to handle China AND everything else, but I would be glad to have China in the game in that way, so I could choose to play there OR in Europe/etc.

It would be better if they did that only for a select few regions, though, as modularity also increases code complexity (and thus maintenance cost). Being able to choose the current map, or a map of China+India (and other adjacent areas), or both together would be optimal IMO.

 

[Andrelvis]

Paradox updates the clawitzengine all the time on the fly. Usually they do it between games though.

Can you name one thing they can add to the engine to make the game run faster? I'm pretty sure they prefer that much more than just 'make game better'.

One thing I can think of is offloading parts of AI-decision making to other threads. For example: AI character decision evaluation could be moved to a separate thread. If the AI character decides to take the decision, then they send a command to the game that is processed in the same way as when players click decision buttons.

There is an RTS game that did this (I can't recall the name ATM, I read an article about it a while ago), but it is certainly not a common feature in strategy games; it was something innovative in their case. How feasible it would be for CK2 in particular depends on the engine's code, and how much change would be necessary.

 

[Serenity84]

They've done a lot of optimization already. After the map was expanded to India things obviously got slower. But they sped up the game later on.

That said, there is only so much you can do. And this is something other Paradox games suffer too. Stellaris for example. Checking so many variables each month just takes its toll.

 

[SvartVind]

Mechanics of Crusader Kings II were designed with a specific game in mind. Also, the game had targets in regards to hardware it had to run on, at the time.
The game had some budget and some time it could be in developement. In the end, it had to sell well enough. And it did. That's why we are here right now, on this forum, 5 years later.

The problem is, the game wasn't designed for all the additional content and mechanics that we have right now. They were added one by one, usually by using mechanics from the base game, redesigning single systems when they could no longer handle what they should handle right now.

Expanding the game out of what it is designed to do is more costly than it would have been to implement it from the scratch.
Any software, including game engines, including CK2 engine, could be modified to serve some other purpose, or to be improved, or to be optimised. The issue is: how much time would that take, and therefore, how much it will cost? And it has to earn the money back, and more - so there's a limit on what's feasible.

And one might think that we'd appreciate a game update focused on optimization and lacking in content... but no. We wouldn't. ;]
The player base as a whole wouldn't. Especially if you want them to pay for it. And, as mentioned before, you don't develop for free.

So, all mechanics, no matter how great they are, have to pass the "Can we do this in CK2, or do we keep it in the backlog for CK3?" test.


A theoritical CK3 would be a different matter entirely, because it would be informed by what went well and what went wrong in CK2.
Many systems that were added on to the previous systems and use mechanics of these previous systems could be redesigned from the scratch, not having to worry about "how do we change the current game that also has thousands of possible combinations of DLC?"

 

[Meneth]

One thing I can think of is offloading parts of AI-decision making to other threads. For example: AI character decision evaluation could be moved to a separate thread. If the AI character decides to take the decision, then they send a command to the game that is processed in the same way as when players click decision buttons.

This is already the case. Only a vanishingly small part of the AI is done serially.

(Also, basically nothing mentioned in this thread has anything to do with the engine; it's all game code)

 

[Andrelvis]

This is already the case. Only a vanishingly small part of the AI is done serially.

That's interesting to hear, thanks for explaining. I suppose that is true for the AI evaluating decisions as well? That seems (but of course, I can't be entirely sure) to be an important performance bottleneck.

 

[Meneth]

That's interesting to hear, thanks for explaining. I suppose that is true for the AI evaluating decisions as well? That seems (but of course, I can't be entirely sure) to be an important performance bottleneck.

Yes, AI evaluation of decisions happens in parallel.
Most script evaluation in general is done in parallel, including both events and decisions. Execution is done in series.

 

[Kitsuka]

What is the engine responsible for?

 

[Foreign Alien]

What is the engine responsible for?

Being an object of hatred for players for no real reason other than because they don't know wtf they're talking about.

 

[Meneth]

What is the engine responsible for?

Some of the things it does:

• Implements or wraps a variety of basic data types, such as strings, fixed point numbers, arrays, etc.
• Serves as a layer between the game and the GPU; takes care of a lot of rendering related tasks
• Handles differences between operating systems
• Handles the basics of parsing and writing to files; basically provides a structure for the game code to parse in a more abstract manner
• Basic structures for handling exchanges of information in multiplayer, as well as the server and client structure
• Provides a variety of UI elements types, such as buttons, lists, windows, etc.
• Implements the launcher
• Crash reporting and similar

Some things that isn't the engine's responsibility:

• Combat
• What the map looks like (including globe vs. flat)
• Provinces
• Multi-threading
• General performance (the main impact of the engine is on graphical performance)
• Diplomacy
• China
• The scripting language, other than its overall structure
• The AI
• Game features

 

[Kitsuka]

So what the main difference between the engine for CK2 and a newer title like HOI4?

 

[Meneth]

So what the main difference between the engine for CK2 and a newer title like HOI4?

More of the same, for the most part.
The newer games have some more standard data types, and are a bit easier to work with to get to do what you want. Some aspects of the graphical processing is a bit faster. Some bugs here and there have been weeded out (example: for a while, on Linux writing and reading 64-bit fixed points didn't actually work properly, so if you had too much gold, saving and then reloading would cause you to end up with negative money). Extra helper stuff for stuff involving the functionality provided by the engine. That kind of thing.

Basically, it's all the same engine, except CK2 was forked from the main branch of development several years earlier. So some of the things done since never made their way back into CK2. But this is generally stuff only of interest to programmers.

Tl;dr: More of the same for the most part.